A New Secret Image Sharing Scheme Based On Compressed Sensing Technology

In traditional secret image sharing schemes,all the data of a secret image have to be processed,which prolongs the algorithm execution.Meanwhile,the inflated data become a burden for network transmission and disk storage when many secret images need to be routinely shared.When the image signal has sparseness or most of the coefficients are zero or approximately zero in the case of a sparse transformation base,compressed sensing technology measures the original image perceptually by constructing a suitable measurement matrix that is not relevant to the sparse base.The measured data cover the vast majority of the useful information of the original image.While ensuring precise reconstruction,the original image is compressed from high dimensional to low dimensional and the amount of image data decreases dramatically.Thus,a number of problems caused by the large amount of data in traditional secret image sharing schemes could be solved by compressed sensing.At the reconstruction end of the signal,an accurate or high-precision approximation of the original image information can be obtained by a certain reconstruction algorithm.In this paper,we combined the traditional secret image sharing with compression sensing technology,and achieved a secret image sharing scheme based on compression sensing technology.In order to further improve the performance of the program,we have optimized our scheme from three aspects:sparse representation of signal,signal perceived measurement and signal reconstruction.Through the experiment we found that our method can clearly reduce the amount of data need to be processed and effectively shorten the algorithm execution time,and the reconstructed images can achieve an ideal accuracy of visual effect.Compared with the Thien-Lin's scheme,the initial scheme achieved in this paper can reduce the average image recovery time by 48.66%and the image revealing time by 29.32%.The optimized solution can reduce the average image recovery time by 58.77%and the revealing time by 58.16%,and the reconstructed image accuracy of the optimization scheme is improved by 4-10dB compared with the initial scheme.